Magnolol derivatives as specific and noncytotoxic inhibitors of breast cancer resistance protein (BCRP/ABCG2)

Abstract

The breast cancer resistance protein (BCRP/ABCG2) transporter mediates the efflux of numerous antineoplastic drugs, playing a central role in multidrug resistance related to cancer. The absence of successful clinical trials using specific ABCG2 inhibitors reveals the urge to identify new compounds to attend this critical demand. In this work, a series of 13 magnolol derivatives was tested as ABCG2 inhibitors. Only two compounds, derivatives 10 and 11, showed partial and complete ABCG2 inhibitory effect, respectively. This inhibition was selective toward ABCG2, since none of the 13 compounds inhibited neither P-glycoprotein nor MRP1. Both inhibitors (10 and 11) were not transported by ABCG2 and demonstrated a low cytotoxic profile even at high concentrations (up to 100 µM). 11 emerged as the most promising compound of the series, considering the ratio between cytotoxicity (IG₅₀) and ABCG2 inhibition potency (IC₅₀), showing a therapeutic ratio (TR) higher than observed for 10 (10.5 versus 1.6, respectively). This derivative showed a substrate-independent and a mixed type of inhibition. The effect of compound 11 on the ABCG2 ATPase activity and thermostability revealed allosteric protein changes. This compound did not affect the expression levels of ABCG2 and increased the binding of the conformational-sensitive antibody 5D3. A docking study showed that 11 did not share the same binding site with ABCG2 substrate mitoxantrone. Finally, 11 could revert the chemoresistance to SN-38 mediated by ABCG2.

Keywords: ABC transporters; ABCG2; Inhibitors; Magnolol; Multidrug resistance.

Figure 1.

Inhibition curves of magnolol derivatives and histogram overlays. IC₅₀ curves of ABCG2 inhibition for 10 (A) and 11 (B). HEK293-ABCG2 cells were exposed to mitoxantrone (5 μM) and magnolol derivatives (10 and 11 at 0.039 – 50 μM). Ko143 (0.5 μM) was used as positive control (100% inhibition). Results were expressed as percent of inhibition related to the positive control. (C) Histograms of ABCG2 inhibition: 10, 11 and Ko143.

Figure 2 -. Cell viability assay and therapeutic ratio representation.

Cytotoxic profile of the two best compounds, 10 (A) and 11 (B). HEK293 and HEK293-ABCG2 cells were submitted at a long-term treatment (72 hours) with increasing concentrations of magnolol derivatives (0.19 – 100 μM). (C) Inhibition potency and cytotoxicity of the compounds with their respective therapeutic ratios.

Figure 3 -

Effects of 11 on ABCG2 ATPase activity and protein thermostability. (A) Effect of 11 at increasing concentrations (0.001 – 50 μM) on basal ABCG2 ATPase activity. Thermostability assay in the absence (B) and the presence (C) of 11. The data represent mean ± SD of three independent experiments performed in duplicate.

Figure 4 –

Mechanism of ABCG2 inhibition caused by compound 11. (A) Intracellular accumulation of Hoechst 33342 (1 μM) in HEK293-ABCG2 cells by confocal microscopy. Ko143 (0.5 μM) was used as control and 11 was tested at 50 μM. (B) Effect of 11 on mRNA expression levels. (C and D) Histograms from the conformational 5D3 antibody shift assay. Overlay of untreated control and cells treated with Ko143 2 μM (C) and 11 50 μM (D). (E) Lineweaver-Burk plot using different concentrations of 11 and mitoxantrone as substrate.

Figure 5 -. Docking of Ko143, magnolol derivatives and mitoxantrone on ABCG2.

(A) Binding site of Ko143 and 11 (spheres representations in purple and orange, respectively) and (B) sequential docking analysis of 10 (in yellow), 11 (in orange) and 12 (in green) on ABCG2 structure (PDB: 6VXI) without mitoxantrone (MTX). (C) Redocked MTX interactions with ABCG2. (D) Binding site comparison of mitoxantrone molecules (original MTX from 6VXI in light blue and redocked MTX in dark blue). (E) Sequential docking analysis and interactions of 10, 11 and 12 (in yellow, orange, and green, respectively) with ABCG2 structure containing MTX (in dark blue). 11 (F) and 12 (G) interactions with ABCG2 in the presence of MTX. (H) Sequential docking and interactions involving two molecules of 11. ABCG2 chains are light gray (A chain) and dark gray (B chain), hydrophobic interactions (C, F, G and H) displayed as dashed gray lines. Amino acids in B, D and E for spatial reference only. Figures were generated using PYMOL (Molecular Graphics System, Version 1.3, Schrödinger, LLC) and PLIP web server [67].

Figure 7 –. Chemosensitization assay using of stably transfected cells overexpressing ABCG2.

Cell viability after 72 h of treatment. HEK293-ABCG2 and HEK293 cells treated with SN-38 alone and HEK293-ABCG2 cells in co-treatment with SN-38 and 11 at saturating concentration or SN-38 and Ko143 at 0.5 μM (positive control). Data represents mean ± SD of at least three independent experiments performed in triplicate. Groups were statistically compared using one-way ANOVA (Sidak’s multiple comparisons test). NS = non-significant; ***p<0.001 and ****p<0.0001.

Scheme 1.

Synthesis of 13. a) dry acetone, K₂CO₃ (2 equiv), MeI (2 equiv), 56 °C, 4h. b) 11 (0.2 M in MeOH), IBX (1.2 equiv), 0 °C, 30 min; Na₂S₂O₄, rt, 5 min.